ARIES Meeting General Atomics, February 25th, 2005

Brad Merrill, Richard MooreFusion Safety Program

Pressurization Accidents in ARIES-CS

Presentation Outline

• Describe the details of a MELCOR model under development for the ARIES-CS dual coolant blanket concept

• Present MELCOR results for reactor startup

• Present the results of a decay heat removal analysis for a complete loss-of-flow-accident (CLOFA) as a benchmarking exercise for our MELCOR model

• Discuss possible accident scenarios that must be considered in assessing the safety of ARIES-CS

• Present results for two vacuum vessel (VV) pressurization accidents with the use of the ARIES-CS MELCOR model

• Conclude by discussing future modeling requirements and analyses

Modeling Assumptions and Parameters

• Modeled one-half a field period segment (1/6th of three field period reactor)– First wall surface area 1/6 of plasma surface area ~ 135 m2

– Applied first wall heat flux of 0.5 MW/m2

– Total operating thermal power for model ~ 372 MW

– Decay heat (obtained from “Initial Activation Assessment for ARIES Compact Stellarator Power Plant” paper ~ 10 MW at shutdown and 3 MW at 60 s)

– Radial build based on Laila’s Nov. 11th radial compositions and an ARIES-ST unit blanket cell

• Helium Loop (~115 m3)– Pressure 8 MPa and inlet /outlet temperature 300/480 °C

– Mass ratio (external /in-vessel) ~ 7 to maintain He velocities below 40 m/s in piping

• PbLi Loop (~255 m3)– Pressure 4 MPa and inlet /outlet temperature 460/700 °C

– Mass ratio (external /in-vessel) ~ 4 (assumed from previous studies)

• Loop configurations based on ARIES-AT confinement building

Header

PbLi drain tank

Heat transfer vault

HXs

Pump

Circ.

MELCOR Model Elevations Based on ARIES-AT

(Need similar information for ARIES-CS)

MELCOR ARIES-CS Model Schematic

CV

020

HS

400

CV

300

CV

400

HS

401

HS

402

HS

403

HS

404

HS

405

HS

450

HS

451

HS

452

CV

401

CV

402

CV

403

CV

404

CV

450

CV

460

CV

320

HS

320

HS

321

HS

322

CV

321

HS

323

HS

324

CV

323

CV

322

CV

324

HS

325

HS

342

CV

310

HS

341

HS

340

fl650

fl651fl245

fl660

fl661fl265 fl275fl240

CV605

CV600

fl28

0

f l65

3

fl65

2

fl66

2

fl66

3

fl30

0

fl23

0

fl64

0

fl63

0

fl20

0

CV610

CV615HS650

HS750CV036

CV630 to CV635

HS22000 to HS22500

CV800 to CV805

CV810HX- Secondary Side

HS

210 0

0 to

HS

2150

0

CV

710

HX

- S

eco

nd

a ry

Sid

e

CV

530

t o C

V5 3

5

CV

700

t o C

V7 0

5

fl710

fl716

HS

520

HS

540

CV

036

CV

520

CV

540

fl41

6

fl410

CV

551

CV

550

fl551

fl430

fl20

5

fl21

5

fl28

1

fl27

1

CV

502

CV

500

CV

503

CV510

fl620

fl600

fl615

fl61

0

fl810fl816 fl290

fl220

fl420

fl425

He toroidal header

He concentric pipe PbLi toroidal header

PbLi concentric pipe

PbLi heat exchanger

He heat exchanger

IB Shield

VV VV

IB blanket OB blanket OB Shield

FW FW

Zone 1 Zone 2

Startup Transient Predictions

Fluid temperatures Structure temperatures

1.000.00 0.25 0.50 0.75 1.00Time (hr)

200

300

400

500

600

700

800

Tem

per

atu

re (

°C)

PbLi outlet temp PbLi inlet temp Helium outlet temp Helium inlet temp

dT/dt ~ 0.8 °C/sPWRs allow ~.008 °C/s => ~ 4 MW to heat loops

0.00 0.25 0.50 0.75Time (hr)

200

300

400

500

600

First wallRibs of blanket

Tem

per

atu

re (

°C)

CLOFA Decay Heat Removal

• Shutdown and complete loss of active cooling flow occurs, with the VV cooling system entering a natural convection mode

• Latent heat re-distribution results in FW temperature of ~ 630 °C (lower than 730°C predicted by Carl, perhaps due to flow coast down)

• By 3 hours the temperatures start to decay

• At ~26 hours the PbLi loop natural convection increases

0

100

200

300

400

500

600

700

800

0 24 48 72 96 120Time (hr)

Tem

per

atu

re (

°C)

First WallPbLiBack WallShieldVacuum Vessel

OB blanket

IB blanket

0 10 20 30 40Time (hr)

-100

-50

0

50

100

150

Mas

s F

low

(kg

/sec

)

Loop

Possible Reference Accidents

• In-vessel helium or water loss-of-coolant accident (LOCA) analysis to demonstrate:

– Pressurization does not fail first confinement barrier (i.e., ARIES-CS vacuum vessel)

– Limited chemical reactions and hydrogen formation (exclude water or limit PbLi)

• In-blanket (breeder zone) helium LOCA analysis to assess:– Blanket and tritium purge gas system pressurization

– Subsequent in-vessel leakage

• Ex-vessel helium and PbLi LOCA analysis to determine:– Pressurization of reactor or HTS vault is tolerable

– Behavior of FW/blanket without active plasma shutdown

• These accidents are the most likely confinement bypass event initiators because of confinement barrier over-pressurization; it must be demonstrated that for all three events the dose at the site boundary does not exceed 10 mSv for all radioactive sources combined (dust, FW oxidation, Po-210, and tritium)

Pressurization of Module Maintenance Vacuum Vessel

• Two pressurization accidents were considered

• A single FW channel rupture (0.0012 m2)

• Helium inlet header (0.04 m2)

• Design basis events with probabilities in the ~ 1x10-3/year range

• Free volume within the vacuum vessel (VV) was set at plasma volume of ~535 m3

• Immediate plasma shutdown occurs, but radiant collapse was not included

Pressurization of Module Maintenance Vacuum Vessel

• Shutdown and loss-of-coolant occurs after 1 hour, and VV cooling enters natural convection mode

• Pressures reach > 18 atmospheres within seconds after the large break while the small break takes ~ 5 minutes to reach 18 atmospheres

• A pressure relief system will be required for this VV option

0

6

12

18

24

0.0 0.5 1.0 1.5 2.0Time (hr)

Pre

ssu

re (

atm

)

Break size = 0.0012 m2

Break size = 0.04 m2

Pressurization of Period Maintenance Vacuum Vessel

• Two pressurization accidents were considered

• A single FW channel rupture (0.0012 m2)

• Helium inlet header (0.04 m2)

• Design basis events with probabilities in the 1x10-3/year range

• Free volume within the vacuum vessel (VV) was set at plasma volume of ~7000 m3

• Immediate plasma shutdown occurs, but radiant collapse was not included

• Shutdown and loss-of-coolant occurs after 1 hour, and VV cooling enters natural convection mode

• Pressures reach ~ 2.0 atmospheres within seconds for the large break while the small break takes ~ 10 minutes to reach 2.0 atmospheres

• A pressure relief system may not be required for this VV option

• However, decay heat removal may be an issue

Pressurization of Period Maintenance Vacuum Vessel

0.0 0.5 1.0 1.5 2.0Time (hr)

0.0

0.6

1.2

1.8

2.4

Pre

ssu

re (

atm

)

Break size = 0.0012 m2

Break size = 0.04 m2

Results Summary

• A decision must be made on which accidents will be selected as Reference Accident Scenarios (pressurization, decay heat removal, and hydrogen generation are concerns)

• Models are being developed to analyze releases during adopted Reference Accident Scenarios, but more design information and benchmarking is required

• Decay heat removal analysis suggests temperatures somewhat less than Carl’s analyses, and this difference must be investigated further

• The VV is the primary confinement barrier and design basis pressurization events must not fail this boundary

– The small VV for the module maintenance scheme will not work without an expansion volume available (will also be part of primary confinement boundary)

– The large VV for the sector maintenance scheme is very attractive as a primary confinement barrier, if the decay heat removal problem can be addressed

Future Analysis Requirements

• MELCOR in-blanket (LOCA) model may have to be developed based on choice of tritium and Po-210 (or Bi-208) extraction systems (low or high pressure extraction systems)

• Tritium inventory, permeation, and accident releases

– Annual release limit (0.1 g-T as HTO/a – ITER to 8 g-T as HTO/a – FIRE) translates into an in-building permeation limit of between ~10 to 800 g-T/a, assuming a 99% efficient tritium cleanup system

– To meet this limit the tritium extraction method proposed by Siegfried Malang, which is a vacuum permeator (Nb tubes), is proposed for the PbLi to lower the T2 concentration to levels (produce pressures above PbLi < 0.2 Pa) where losses to the building are below the annual release limit

– Will need tritium production rates

• A TMAP model should be developed to examine tritium permeation in the blanket and from the cooling systems, and to estimate tritium accident mobilization rates

• Need estimates of VV PFC erosion dust and FW activation product inventories

• ARIES-AT PFC dust and Po-210 inventories will be used for accident source terms

• This information will feed into MELCOR LOCA and LOVA predictions for radioactive release estimates and site boundary dose predictions